go thinner go wider

NVIDIA G80 Chip Details


Main Chip Details
Manufacturer NVIDIA
Process 90nm
Foundry TSMC
Transistor Count 681M
Die Size 484mm²
21.5mm x 22.5mm
Chip Package Flipchip + HS
Memory Bus
Bus Width 384-bit
Memory Channels 6 x 64-bit
Memory Type Support GDDR to GDDR4
Geometry Pipeline
Geometry Processing VS4.0
Processor Configuration 128 VS
Rasteriser
Basic Pipeline Config 32 / 24 / 192
Textures / Pixels / Z
Textures Per Pass 128
Basis Filtering Bilinear
Texture Filtering Methods Bilinear, Trilinear, 2-16x Anisotropic
Texture Compression DXTC 1-5, 3Dc+
Fragment Processing PS4.0
Fragment Processing Precision FP32
Fragment Processors 128x scalar MADD+MUL
FSAA
FSAA Support Multisampling, Supersampling, Coverage sampling, Transparency
ROP Subsamples 4
Subsampling Method 2x/4x/8x RGMS
Further Details
API Compliance DX10.0
System Interconnect PCI Express x16
Display Pipeline None (NVIO)
Multi GPU Support Multi cascade

Notes

First consumer Shader Model 4.0 GPU, fully scalar, unified, hardware load balanced

NVIDIA G80-based Boards

NVIDIA GeForce 8800 GTX 768MiB
NVIDIA GeForce 8800 GTS 640MiB
NVIDIA GeForce 8800 GTS 320MiB
NVIDIA GeForce 8800 Ultra 768MiB